/evaluators

AEO · aeo-citation-readiness

Source Attribution Evaluator

Checks whether important claims are backed by source names, dates, methodology, or first-party evidence that AI systems can cite.

What it checks

AI engines are less likely to cite unsupported claims. Named sources, methodology pages, and original data make answers more defensible.

Pass criteria

  • Statistics include a source and year.
  • Product claims link to docs, changelog, methodology, or customer proof.
  • The page distinguishes measured data from opinion or projection.

Example

According to AEO Pro benchmark data collected on 2026-05-20, pages with visible FAQ schema and llms.txt passed 2.1x more AI citation-readiness checks than pages without either asset.

Common failures

  • - Naked statistics without a source.
  • - Testimonials with no role, company, or context.
  • - Claims that appear only in screenshots or images.

Related surfaces

FAQ

Do all claims need external sources?
No. First-party product claims can link to your own docs, changelog, methodology, or benchmark data. The important part is making the evidence explicit.